Twitter Malicious Account and Content Detection using Machine Learning

Abstract

The capacity to link individuals all over the globe to share media and ideas has contributed to the meteoric rise in popularity of social networks. Making false accounts is a big problem on these networks, and one that consumers are really worried about. As an example, Twitter permits an excessive quantity of spam because it has grown into one of the most widely utilized platforms ever. False accounts promote unwanted services or websites through tweets, which impacts legitimate users and disrupts resource utilization. A popular field of research in current online social networks is the detection of spammers and the identification of fraudulent users on Twitter. Recent polls conducted by research delegates included a taxonomy that classifies Twitter spammers into two broad categories. The first group deals with the topic of detecting fake content, which is often accomplished through the use of a regression prediction model, or lfun scheme. Fake User Identification is another subcategory that focuses on identifying fraudulent Twitter users using a combination of user- and content-based criteria.